Endogenous cortisol production in response to knee arthroscopy and total knee arthroplasty

There is controversy about whether patients who take exogenous glucocorticoids, such as prednisone, require supplemental (exogenous) glucocorticoids in order to meet the physiological demands of surgery. In this study, we sought to define the magnitude of the surgical stress response in normal patients undergoing major and minor elective orthopaedic surgery.

A prospective, observational study of thirty patients who had not taken exogenous glucocorticoids and who underwent either elective knee arthroscopy or elective unilateral total knee arthroplasty was performed. Regional anesthesia was used for all patients, and all patients treated with total knee arthroplasty had continuous epidural anesthesia for forty-eight hours after the surgery. The stress response was assessed on the basis of serum and twenty-four-hour urine cortisol levels; comparisons of the urine values were made after correcting for renal function by calculating the cortisol-to-creatinine clearance ratio.

Preoperatively, patients undergoing arthroscopy and total knee arthroplasty had similar cortisol-to-creatinine clearance ratios. Patients treated with total knee arthroplasty had a significant (p < 0.001) surgical stress response on the day of the surgery, compared with baseline, whereas patients treated with arthroscopy did not. The mean cortisol-to-creatinine clearance ratio in patients treated with total knee arthroplasty was highest on the day of the surgery and decreased on the third postoperative day. However, on the third postoperative day, the cortisol-to-creatinine clearance ratio still was significantly higher than the baseline value (p < 0.001). Significant differences in the serum cortisol levels also were detected between the patients treated with arthroscopy and those treated with total knee replacement.

Patients undergoing total knee arthroplasty had a significant surgical stress response (a seventeenfold increase in the cortisol-to-creatinine clearance ratio); patients treated with arthroscopy did not. Additional studies, including a prospective trial of patients taking exogenous glucocorticoids, are warranted. Until they are performed, the significantly increased cortisol production observed in non-steroid-dependent patients following total knee arthroplasty leaves open the possibility that steroid-dependent patients undergoing this procedure could benefit from perioperative glucocorticoid supplementation. Since the non-steroid-dependent patients in the present series did not mount a substantial stress response to knee arthroscopy, our results do not support the use of supplemental steroids for that less-invasive procedure.